Illiotibial band syndrome (ITBS) is a common disorder amongst recreational and competitive runners. In fact, it ranks only behind patellofemoral pain syndrome in frequency(1). The iliotibial band (ITB), a tough band of connective tissue that runs from the hip to the lateral knee, serves as the attachment for the gluteus maximus and the tensor fascia... MORE
How reliable is the gait lab in your pocket?
Patellofemoral pain (PFP) plagues up to 17% of runners who complain of knee pain(1). Often weak gluteus medius muscles are to blame with resulting hip adduction (HADD) and positive Trendelenburg sign during the unilateral stance phase of running. Patellofemoral pain also affects knee flexion (KFLEX) in single stance, with runners avoiding KFLEX to prevent pain. The two issues – weak hip abductors and flexion avoidance – may result in knee valgus, which includes HADD, tibial adduction, and internal femoral rotation.
Clinicians target their therapeutic program to amend these issues, relieve pain, and return the runner to sport. Assessing the progress of the program is often based on the subjective observation of running gait. Ready access to videography on smartphones and tablets now enables clinicians to record running pre and post-intervention and use the video as an outcome measure.
However, is the snapshot video taken in the clinic or field as accurate and reliable as a three-dimensional (3-D) gait lab analysis? Researchers in the UK hoped that, indeed, clinicians could use a smartphone to gather markerless high-frame two-dimensional (2-D) video of runners with PFP and gather reliable data that compared favorably with a 3-D lab. They recruited 21 subjects (10 males and 11 females) with PFP to participate in a study comparing the two methods of gait analysis.
The study was conducted in a gait lab with a four-camera infrared motion analysis system. Each subject was marked at the typical bony landmarks, the midpoint of the thigh and leg, and on the outside of the shoe. In addition to the gait lab cameras, the investigators placed a smartphone camera ahead of the runner to capture them running toward the camera, and beside the ground-embedded force plate to film the sagittal view when the runner struck the force plate. Each smartphone recorded at a speed of 240 frames per second as individuals ran the 13-meter track stepping on the force plate with their affected leg roughly halfway through the run.
The data from the gait lab was analyzed using a customized program to measure peak HADD and KFLEX. The video from the smartphones was loaded into the Hudl Technique application on a tablet device(2). Researchers used the angle finder within the Hudl Technique app to mark anatomical landmarks and measure peak HADD, contralateral pelvic drop, and KFLEX (see figure 1). There was moderate intra-rater reliability for peak HADD and KFLEX using the tablet and the Hudl technique. However, the inter-rater reliability was poor, meaning the chance of two people getting the same results from using the same image was small. This variability calls into question the reliability of using a smartphone camera and associated app to measure kinematics in a runner.
Figure 1: Hudle Technique Application
Meanwhile, the comparison between the 3-D and 2-D findings was also poor. Neither of the 2-D measures was equal to the 3-D results. The mean 2-D measures of peak KFLEX were significantly off from the 3-D ones.
While 3-D gait lab analysis is the gold standard of kinematic study, it isn’t practical for most clinicians. The researchers were hoping to find an easy and affordable way to bring an element of the gait lab to the practitioner. Previous studies have reported the successful use of smartphone cameras in gait analysis but used different software. The Dartfish software seems to offer more precision in that users review the video on a computer screen that is much larger than a tablet(3). It also uses a mouse to mark bony landmarks rather than a finger, which is less precise. The downside to software over applications is the cost.
The other aspect lacking in the 2-D analysis is the absence of imaging in the transverse plane. That said, the authors agree that 2-D video may still be useful. They suggest placing the camera as close to the subject as possible on a treadmill, rather than over six meters away as in this study. They suspect that the distortion in their video came from the distance rather than the frame rate, and affected the accuracy of the Hudle app. They also recommend using retroreflective markers when filming to make it easier to identify bony landmarks in the video.
It appears that when obtaining video to measure the effectiveness of an intervention, any smartphone camera will do. It’s the marking and measuring of the video that needs improved accuracy. Therefore, if the budget allows, invest in software that analyzes the video on a bigger screen with precision marking.
- Physical Therapy in Sport. 2020;43:36-42
- Hudl, Agile Sports Technologies Inc, Nebraska, USA
- Dartfish, Fribourg, Switzerland